Bengaluru, India

Dinesh Mittal

Engineering leader · Platforms, Cloud and AI

I write about building and running AI and platform infrastructure that is fast, reliable and affordable, from 26 years of doing it.

About

I have spent 26 years in engineering, from 3G protocol stacks and Wi-Fi firmware to cloud platforms and AI infrastructure, at CSR/Qualcomm, Intel, OLA Krutrim and Palo Alto Networks.

I have led organisations of 120+ people across India, Europe, the US and East Asia, took an AI cloud from zero to public launch, and ran the platforms behind more than $6B a year in revenue. Today I am a co-founder of NeoSmith AI, where we build codebase-specific small language models and a routing layer that sends each request to the cheapest model able to answer it.

I write here about AI infrastructure, platform engineering and running engineering teams: what worked, what it cost, and what I would do differently.

More about me

years in engineering
26
people led across 4 regions
120+
yearly revenue on my platforms
$6B+
yearly efficiency delivered
$10M+

NeoSmith AI · Palo Alto Networks · OLA Krutrim · Intel · CSR (acquired by Qualcomm) · L&T Infotech

Writing

Latest essay · 23 Sept 2026 · 2 min read

Why my notes and my website live in two separate repos

One private vault for everything I think, one site for what I choose to share, and a small script as the only bridge between them.

Read the essay

Experience

  1. Jun 2026 – Present

    Co-founder, Strategy, Product and Engineering · NeoSmith AI

    Codebase-specific small language models behind a routing layer that sends each request to the cheapest model able to answer it. MVP in two months; enterprise pilots show AI coding spend down by more than 80% against frontier-model usage.

    • Small language models
    • Inference routing
    • AI economics
  2. May 2025 – Jun 2026

    Senior Director, Head of Infrastructure Platform Engineering · Palo Alto Networks

    Led the platform engineering group behind one-touch networking, compute and identity services for internal engineering. An agentic AI platform for DevOps and SRE work cut P1/P2 incidents and mean time to detect by 80% each, at more than 99.5% uptime.

    • Platform engineering
    • Agentic AI
    • SRE
  3. Mar 2024 – Apr 2025

    Head of Engineering, Cloud Platform · OLA Krutrim

    Built the team and engineering practice from scratch and took a general-purpose and AI cloud from zero to public launch in six months: compute, storage, networking, security and Gen-AI model services. Moved OLA group workloads off hyperscalers and grew enterprise revenue to $30M a year.

    • AI cloud
    • Cloud platform
    • 0 to 1
  4. Aug 2013 – Feb 2024

    Senior Director / Director of Engineering · Intel

    Led DevOps and infrastructure engineering for the AI, Wireless, IoT, Network and Edge businesses, serving Apple and hundreds of OEM and ODM partners on platforms behind more than $6B a year in revenue. The same headcount delivered 4x the volume at 4x the velocity, and validation automation rose from under 20% to over 80%.

    • DevOps
    • Infrastructure
    • Validation at scale
  5. Jun 2001 – Aug 2013

    Senior Engineering Manager, Principal Engineer and Architect · CSR (acquired by Qualcomm)

    Built and led Wi-Fi systems, integration, validation and certification; more than 100 million units of wireless software shipped. In-house certification labs and chambers saved more than $2M a year. Earlier, architected 2G/3G/4G protocol stacks and Bluetooth host software.

    • Wireless
    • Embedded
    • Protocol stacks
  6. Nov 1999 – May 2001

    Software Engineer, contracted to Samsung Mobile · L&T Infotech

    Built the RRC layer of a 3G protocol stack.

    • 3G

How I work

  1. 01

    Cost is a design input

    Unit economics belong in the architecture review, not in the post-launch surprise. I ask what a request costs before asking how fast it is.

  2. 02

    From silicon to service

    Having worked from embedded systems to cloud platforms, I look for the layer where a problem is cheapest to fix, which is often not the layer where it shows up.

  3. 03

    Platforms are products

    Internal platforms succeed when engineers choose them. That means clear interfaces, honest documentation, and measuring adoption instead of mandating it.

  4. 04

    Small teams, clear ownership

    Most delivery problems are ownership problems. I prefer fewer, smaller teams with explicit boundaries over large groups with shared responsibility.